Developing SonReb models to predict the compressive strength of concrete using different percentage of recycled brick aggregate

Abstract

This study was conducted to determine the relationship between nondestructive and destructive tests in concrete cubes using different ratios of normal stone and recycled brick as coarse aggregates. Variations in the grade of concrete, density, and age were considered to make the model prediction more efficient in places where the use of recycled brick aggregates is common. Normal concrete grades M20, M25, and M30 were considered having density variation by replacing stone with recycled brick aggregate, and age by testing concrete strength at 7, 28, and 84 days. A regression model was created using artificial neural networks and multiple regression analyses. The study showed that the regression model developed using an artificial neural network predicted better results. The models obtained from the experiment were compared with other models provided by different authors. The study also considered the effect of using recycled brick aggregate in nondestructive tests and the modulus of concrete.